Railway-built image pull fails in EU West after successful push
jzoepc
OP

14 days ago

Hello Railway team,

Our GitHub-backed Python/Django web service in EU West (ams) cannot deploy an update. The build succeeds and the build logs show the image export and image push completing, but deployment fails at "Deploy > Create container" before the application starts.

The error says the built image on production-dpams.railway-registry.com "could not be pulled from the registry" and suggests checking private registry credentials. This is an image built by Railway from our GitHub repository, rather than an image from a separately configured private registry.

This happened twice on October 28, 2026: the automatic deployment around 22:44 EDT, and one manual redeploy around 22:45 EDT using the same configuration. Both failed at the same step. The latest retry used Railpack 0.43.0 and Python 3.13.16; its build log shows the 176.3 MB image push completing.

Railway's built-in diagnosis reports "Infrastructure Error" and recommends contacting support if a redeploy fails again. Our previous deployment remains active and the live site responds, so the current impact is that we cannot release the update, rather than a complete outage.

Could you investigate whether the generated image is available to the runtime in ams and whether there is a registry authorization or replication problem? Please advise on a safe recovery that preserves the currently active deployment. The affected service is selected on this thread. I have both failed deployment IDs and can provide them through an appropriate private channel if needed.

Thank you,

Jefferson

Solved

1 Replies

Status changed to Awaiting Railway Response Railway • 14 days ago


dizzydes90
EMPLOYEE

14 days ago

We looked at both deployments from October 29 at 02:44 UTC and 02:45 UTC (22:44 and 22:45 EDT). In both, the build finished and the image was pushed. Each one then failed when the image was pulled onto the deploy host in EU West. You don't need to set up any registry credentials for an image we built from your GitHub repo, so that hint in the error doesn't apply to your service.

A failed deployment never replaces the active one, so your current deployment keeps serving traffic. A failed deploy can't take the live site down.

To retry safely, start a fresh deployment rather than redeploying one of the failed ones. Open the service, press Cmd+K (Ctrl+K on Windows) and choose Deploy Latest Commit, or push a new commit. Either way the image is built and pushed again from scratch, and your previous deployment stays live until the new one is healthy.


Status changed to Awaiting User Response dizzydes90 • 14 days ago


Railway
BOT

7 days ago

This thread has been marked as solved automatically due to a lack of recent activity. Please re-open this thread or create a new one if you require further assistance. Thank you!

Status changed to Solved Railway • 7 days ago


Welcome!

Sign in to your Railway account to join the conversation.

Loading...